This is malpractice. Clinton National Field Director Guy Cecil’s January 19 internal memo discussing February 5th’s congressional districts and the threshold numbers from gaining or defending the gain of an extra delegate is replete with error.
Set aside the malpractice of discontinuing polling in caucus states where the blind-flying Clinton campaign allowed Obama’s team to run up the score, this revelation shows that the Clinton’s HQ apparently did not have simple calculators.
With proportional allocation, since delegates are rounded up and run to the thousandth decimal, to gain a 4-2 split on needs to win 3.5/6ths of the vote. In 6-delegate districts, that percentage is 58.334%. This is undoubtedly what Cecil is referring to when he cites “59%” as the blanket significant threshold. In 4-delegate districts, 2.5/4ths is 62.500%. In 8-delegate districts, the 5-3 split number is 56.250% (4.5/8ths). In 7-delegate districts, a 4-3 becomes 5-2 at 64.286% (4.5/7ths).
Cecil specifies a 59% threshold for 22 critical run-up Clinton's or hold-down Obama's score districts – 16 strong Clinton districts and 6 strong Obama districts. The strong Clinton districts, according to the internal memo, were AL-6, AZ-7, NJ-16, NJ-17, and CA districts 18, 19, 21, 23, 31, 32, 34, 38, 39, 43, 45, and 51. (For those of you wondering about New Jersey’s nonexistent 16th and 17th congressional districts, New Jersey structured delegate allocation by “delegate districts,” of which there were 20.)
Not one of Clinton's 16 favorable districts that Cecil cites were 6-delegate districts. In 14 of them, there were four delegates, the extra-delegate threshold for a 3-1 split being 62.500%. New Jersey’s DD-16 was only 3 delegates, meaning a bare one-vote majority up to a 66.666% margin of victory would result in a 2-1 split. Spending effort running up the score in 3-delegate districts that are comfortably yours by majority simply doesn’t get any more wasteful and Clinton came nowhere near the 3-0 shutout. Finally, Clinton-favored CA-23 was a 5-delegate district where a bare majority gets the winner 3-2, but a 40% win is required to get a 4-1 split. Obama won it outright.
In the six Obama-favored districts where Cecil advocated playing defense (AL-1, AL-2, GA-3, GA-4, GA-5 and TN-8), three of them (AL-1, AL-2, GA-3) were 4-delegate districts, TN-8 was a 5-delegate district, and GA-5 was a 7-delegate district. Out of all 22 districts where Cecil cited “59%” as the critical threshold, only one – GA-4 – was actually such a district. Obama won it, with 79.441% of the vote, and got a 5-1 split.
In Clinton’s defense, she was actually successful in California in most targeted districts with a 3-1 split, with extra effort probably making a difference in six districts where she won between 62.500% and 65% of the vote and might not have gained the extra delegate without it. On the other hand, she won five California districts so comfortably (with 69%-76% of the vote and victory margins of between 38%-52%) that worry about 58.334% and a 16.667% margin of victory was probably moot to begin with. Had Clinton spent her 38%-52% winning margin districts in 5-, 6-, and 7-delegate districts, she'd have been far more efficient racking up delegates. Her campaign did not understand this.
Failure to understand the math clearly hurt in a few other places. For example, in AZ-07, Clinton eclipsed the 58.334% threshold with 58.501% of the vote… and the delegates split 2-2. In CA-19, Clinton eclipsed 58.334% with 59.965%… and the delegates split 2-2.
The upshot is that the Clinton camp missed easy preparations and unnecessarily wasted and/or misdirected valuable effort. With limited post-Iowa resources, they made miscalculations that surely gave away free delegate points at a time when the pledged delegate race was very tight. With the nomination now long-settled and many other mistakes pointed out by others, the purpose is not to belabor what went wrong because assuredly many things did. Simply, as a site dedicated to the efficient and accurate use of data by political campaigns, we could not let this revelation from Josh Green's piece pass without comment.
At this point, it’s much more a cautionary tale for future campaigns to make sure they hire people who know how to work a calculator and look up some basic information. High school interns would probably do it for free. In short, one key aspect of the epitaph on Clinton’s 2008 campaign will be that simple numbers that any old math-minded person could figure out escaped her top people.